Abstract: A variable interval fuzzy quantification algorithm with self-adjustable factor in full
domain is proposed in this paper. It focuses on digital inverted plasma arc cutting power and studies
strong nonlinearity and uncertainty of power. The neural network is also introduced to decouple
cutting parameters variables in the multi-parameters coupling cutting process. This algorithm avoids
complex nonlinear system modeling and realizes real-time and effective online control of cutting
process by combining advantages of fuzzy control and neural network control. Furthermore, the
optimized fuzzy control improves steady-state precision and dynamic performance of system
simultaneously. The experimental result shows that this control improves precision, ripples, finish
and other comprehensive index of work piece cut, and plasma arc cutting power supply based on
fuzzy-neural network has excellent control performance.
Abstract: B-spline method plays an important role in data interpolating, fitting and smoothing.
Through analyzing the computed path of the de Boor-Cox algorithm, a new method for calculating
B-spline function is given based on the vector extending operation. The proposed method has a
parallel computing architecture and is suitable to realize by computers. Compared with the de
Boor-Cox algorithm, the proposed method can improve computing efficiency by (2k+1) times in the
task of calculating all nonzero values of k-degree B-spine functions. As an application in B-spline
curve fitting, it shows that the proposed algorithm can reduce the computing efforts obviously.
Abstract: Aiming at hole filling in points cloud data reconstruction, a novel neural network
arithmetic was employed in abridged points cloud data surface reconstruction. Radial basis function
neural network and simulated annealing arithmetic was combined. Global optimization feature of
simulated annealing was employed to adjust the network weights, the arithmetic can keep the
network from getting into local minimum. MATLAB program was compiled, experiments on
abridged points cloud data have been done employing this arithmetic, the result shows that this
arithmetic can efficiently approach the surface with 10-4 mm error precision, and also the learning
speed is quick and hole filling algorithm is successful and the reconstruction surface is smooth.
Different methods have been employed to do surface reconstruction in comparison, the results
illustrate the error employed algorithmic proposed in the paper is little and converge speed is quick.
Abstract: A scheduling approach using genetic algorithms (GA) was presented to optimize multiple
projects for quality project period performance with resource constraints. The model of the approach
and key parameters of the algorithm including chromosome encoding and decoding, fitness
computation, initial population, selection and crossover were conducted. A precedence feasible list
was used in the chromosome encoding and decoding operation to reduce search space. An efficient
crossover method was developed to avoid the procedure of chromosome recovery. A comparison was
made between the algorithm and a heuristic scheduling method with an example. The result validates
the superiority of the approach.
Abstract: There is a magnetic field inside thrust bearing. The magnetic field distribution is one of
the key factors which decide the running state and performance. With the finite element of ANSYS
software, the magnetic field of thrust bearing has been analyzed. The rule of changes between the
magnetic field and the influencing factors has been discovered. The analysis shows that the
direction of exciting current should be opposite other than identical. The greater the ratio of gap
length between stator and rotor and working gap length is, the less magnetic field leakage is. The
flux leakage of thrust bearing attains 37 percent when the ratio of the gap is 4.0. If the design of a
complete machine is not proper, the flux leakage may attain 20 percent upwards through other gaps
although that of the bearing itself is almost zero. Calculation of bearing capacity indicates that
theoretical value is not exact due to its neglect of magnetic saturation and flux leakage, while FEM
value comparably matches the real substance with considering the magnetic saturation and flux
leakage. The research provides the theoretical basis for optimization of system structure and
selection of current direction control.
Abstract: Electric steering engine was generally applied in arm systems such as carrier rocket and
missile, its life period is very short, and the space and weight of its reducer were strictly restricted,
so it is not fit to adopt the conventional rule of infinite life to design electric steering engine gear
reducer. This paper carried out the optimum design for it based on the rule of finite life and space,
provided and demonstrated a new design scheme, and realized the interior overall structure design
of reducer. We used the UG plane to realize three-dimension module design and virtual assembly of
gear reducer, elaborated the implementation method of three-dimension motion simulation by the
parameter drive mechanism, and simulated the actual working procedure. Experimental results
indicated that this gear reducer satisfied excellently the requirement of big drive ratio under the
finite life and space.
Abstract: Expected path of polishing tool is one of the most essential needs for movement
scheduling and movement controlling of polishing robot in free-form surfaces polishing. By
analyzing the expected movement and position of polishing tool and based on the traditional
movement scheduling methods, this paper carries out systematic research works on
contour-parallel-machining tool path planning method and direction-parallel-machining tool path
planning method for polishing tool paths figuring out. Compared with contour-parallel-machining
tool path planning method, the direction-parallel-machining tool path planning method needs one
less number of degree of freedom and is much easier to avoid physical interventions and mechanic
singularity, so it is an improved one.
Abstract: This paper studies the problem of motion planning of wheeled mobile robot (WMR),
which is driven by double wheels separately. It mainly accomplish the motion planning of WMR by
taking center of one of double wheels as a base point, and also carry out this problem by traditional
method, which takes the midpoint of interval between two wheels as a base point. Besides, these two
motion planning methods are compared. Experimental results show that the first motion planning
method is simply controlled, saves much computer time and it establishes the foundation to solve the
problem of real time control to WMR.
Abstract: The crux of injection mold production is scheduling, which should be made before the
mold manufacturing. Generally, a mold part has some operations subject to certain sequence
constraints. Furthermore, some operations have to satisfy process combination constraints with one
or more other operations. The expression method of operation constraints is studied. If genetic
algorithm is employed to solve scheduling problems in injection mold manufacturing, invalid
chromosome representation during the computing period of population initialization and the
crossover, mutation will generated unavoidably. An adjustment algorithm is proposed to ensure the
validity of chromosome representation. Any chromosome representation based on operation or
workpiece could be adjusted to a valid operation sequence by this algorithm. The feasibility of the
algorithm is validated by computing cases.
Abstract: The architecture and operation theory of Giant Magnetostrictive Accurate-motion
Actuator have been introduced. After analysing the driving characteristic of giant magnetostrictive
material and requirement of driving power, a design method of wide range and high precision NC
constant-current source has been put out. The output circuit is composed of serial 12-bit DACs
Max531, low-noise high-speed precision operational amplifiers OP27 and driving circuit. It
provides current from 0 to 2.048A with 0.5mA step value. Two fully differential input channels
16-bit, sigma-delta ADCs AD7705 collects output current in feedback loop. Current ripple is
controlled under 0.25mA through using homemade high-performance linear power. The result
shows that the driving power with characteristic of high stability and fast response meets the needs
of driving of Giant Magnetostrictive Accurate-motion.